H2C: Hippocampal Circuit-inspired Continual Learning for Lifelong Trajectory Prediction in Autonomous Driving

2025-08-01

AI Summary

Catastrophic forgetting severely hinders lifelong trajectory prediction in autonomous driving, where models must continuously adapt to evolving traffic scenarios without forgetting previously learned knowledge. Method: This paper proposes Hippocampus-inspired Hippocampal-to-Cortical (H2C), a continual learning framework inspired by hippocampal memory replay mechanisms. H2C employs a dual-strategy sample selection scheme—balancing diversity and distribution representativeness—to automatically construct high-quality memory buffers without task labels or human-annotated distribution shift information. It further introduces a memory replay loss to enforce knowledge retention during incremental learning. Contribution/Results: H2C significantly improves generalization stability across dynamic, real-world driving scenarios. On the INTERACTION benchmark with multiple heterogeneous scenarios, it reduces catastrophic forgetting by 22.71% on average over baseline continual learning methods, demonstrating strong efficacy and practical deployability in real-world autonomous driving systems.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Search and Optimization: Learning to Search
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Deep learning (DL) has shown state-of-the-art performance in trajectory prediction, which is critical to safe navigation in autonomous driving (AD). However, most DL-based methods suffer from catastrophic forgetting, where adapting to a new distribution may cause significant performance degradation in previously learned ones. Such inability to retain learned knowledge limits their applicability in the real world, where AD systems need to operate across varying scenarios with dynamic distributions. As revealed by neuroscience, the hippocampal circuit plays a crucial role in memory replay, effectively reconstructing learned knowledge based on limited resources. Inspired by this, we propose a hippocampal circuit-inspired continual learning method (H2C) for trajectory prediction across varying scenarios. H2C retains prior knowledge by selectively recalling a small subset of learned samples. First, two complementary strategies are developed to select the subset to represent learned knowledge. Specifically, one strategy maximizes inter-sample diversity to represent the distinctive knowledge, and the other estimates the overall knowledge by equiprobable sampling. Then, H2C updates via a memory replay loss function calculated by these selected samples to retain knowledge while learning new data. Experiments based on various scenarios from the INTERACTION dataset are designed to evaluate H2C. Experimental results show that H2C reduces catastrophic forgetting of DL baselines by 22.71% on average in a task-free manner, without relying on manually informed distributional shifts. The implementation is available at https://github.com/BIT-Jack/H2C-lifelong.

Problem

• Addresses catastrophic forgetting in deep learning for trajectory prediction
• Enhances lifelong learning in autonomous driving across varying scenarios
• Uses hippocampal-inspired memory replay to retain prior knowledge

Innovation

• Hippocampal circuit-inspired continual learning method
• Selective recall of learned samples for knowledge retention
• Memory replay loss function for updating knowledge